google-research / mint

Multi-modal Content Creation Model Training Infrastructure including the FACT model (AI Choreographer) implementation.
Apache License 2.0
497 stars 85 forks source link

About the eval result #53

Open nesciemus opened 2 years ago

nesciemus commented 2 years ago

my eval result using the given checkpoint has only a few second valid dance. after a few second, the result became totally nonsence. is this the real case or did I do something wrong?

CuberFan commented 2 years ago

me too

nesciemus commented 2 years ago

me too

so that's the real case and they only shows some beatiful results selectively?

CuberFan commented 2 years ago

me too

so that's the real case and they only shows some beatiful results selectively?

maybe, you can see this issue: https://github.com/google-research/mint/issues/42

nesciemus commented 2 years ago

me too

so that's the real case and they only shows some beatiful results selectively?

maybe, you can see this issue: #42

ok, seems that i should retrain the model from scratch ... Thanks a lot for the reply.